[{"_1":2,"_3":-5,"_4":-5},"loaderData",{"_5":6,"_7":8},"actionData","errors","root",{"_28":-5,"_38":-5,"_39":40,"_41":42,"_43":44,"_45":46,"_47":48,"_49":50,"_51":52,"_53":54,"_55":56,"_57":58},"routes/_store+/_readers+/email_author.$slug/route",{"_9":10,"_11":-5},"book",{"_12":13,"_14":15,"_16":17,"_18":19,"_20":21,"_22":23},"currentUser","id","Z2lkOi8vbGVhbnB1Yi9TdG9yZTo6Qm9va1Byb2R1Y3QvODAyNzg","title","From Flood to Cause","state","published","contactAuthorMethod","email","aboutText","

Something trips. In the next four seconds, two hundred alarms fire. The board is a wall of red, every annunciator is screaming, and the one alarm that actually matters — the cause — is buried somewhere in the flood, indistinguishable from the two hundred consequences stacked on top of it. \"Loudest alarm wins\" is how that ends badly.

And now there's a new temptation: point a large language model at the flood and ask it what happened. It will give you a fluent, confident, beautifully-worded answer — and you will have no way to know whether any of it is true.

From Flood to Cause is written to resolve both problems at once, around a single governing principle:

The engine decides, the model explains.

A deterministic causal engine — not a neural network — identifies the root cause from an engineered fault-propagation graph and corroborating telemetry. The language model never diagnoses, never ranks, never votes. It runs strictly downstream, in an advisory-only role, to put the engine's conclusion into cited language a person can audit — and it abstains the moment grounding is absent. This book shows you, end to end, how to build a system like that and why every piece is where it is.

Try the Live Console Now!

This is a hands-on companion to NEXUS-1, a fully interactive, browser-based root-cause console. No server, no installation, zero setup. Every incident, graph node, edge weight, and delay window in the book is drawn from this live system — nothing is invented. Load it, trigger a cascade, and watch the engine collapse the flood to a single origin:

👉 Play with the live console: https://gregory82gr.github.io/Nexus-1-phase-0/

What's Under the Hood? (The Engineering Blueprint)

This is not a prompt-engineering listicle. It's the literal architecture and the discipline that makes an LLM safe to put near an operating decision:

The Engineered Causal Graph — backbone, learned-weight, artefact, and rejected edges; delay windows; how the engine separates the one origin from the cascade it caused.
The Grounding Triangle — database, telemetry, and documentation as three independent witnesses, assembled into a single Unified Context.
The Anti-Hallucination Engine (H1–H10) — ten concrete controls, each with a worked example: the closed-world contract, the grounding gate, entity and citation validation, structured output, forced abstention, and a hash-chained audit trail.
What the Model May Read — building a RAG corpus you can trust: hybrid retrieval, retrieval evaluation, and corpus governance.
The Reference Stack — OpenTelemetry, Prometheus, Grafana, and Microsoft Semantic Kernel used as a constrained orchestration layer (explicitly not an agent that picks its own tools), implemented in C# and SQL Server 2025 with vector retrieval.
Three Case Studies from the Console — a cascade, an EMI artefact, and a QA escape, each diagnosed and explained from real console data.
Five Developer Appendices (F–J) — copy-pasteable quickstart: project layout, the Ollama Modelfile, the closed-world prompt and validation schema, observability wiring, and a CI evaluation harness — everything needed to stand the explain side up on your own machine.

How Every Chapter Is Structured

Every chapter follows the same engineering blueprint:

What You Already Know — the bridge from your existing software, data, or control-systems background to the concept at hand.
The Concept — a clear, prose explanation of the mechanism and why it exists.
One Figure — a single precise diagram, drawn from real console data, never decoration.
See It in NEXUS-1 — exactly where that concept lives on the live console screen.
Honest Boundary — a transparent statement of what the Phase-0 system genuinely does versus where it simplifies, including the limits of the controls themselves. No claim that something works better than it does.

👤 Who This Is For

If you are a software, data, control-systems, or SRE engineer — comfortable with C#/.NET, SQL, and systems design — and you are being asked to put a language model somewhere its answers actually matter, this book is built for you. It assumes no background in alarm management or nuclear engineering; it builds the reasoning up from the screens and the code. It is the diagnostic-and-AI companion to From Grid to Core, and it stands entirely on its own.

Author: Grigorios Kyriakos Agathangelidis
Greek name: Γρηγόριος Κυριάκος Αγαθαγγελίδης
Also searchable as: Αγαθαγγελίδης Γρηγόριος, Αγαθαγγελιδης Γρηγοριος, Grigorios Agathangelidis.

","authors",[24],{"_12":25,"_26":27,"_28":29},"Z2lkOi8vbGVhbnB1Yi9TdG9yZTo6UHJvZHVjdEF1dGhvci84NTg0MA","name","Grigorios Agathangelidis","user",{"_30":31},"avatar",{"_32":33,"_34":35,"_36":37},"url","https://d39qdlcrvnra4b.cloudfront.net/avatars/3737581/medium/avatar.jpg?1783850356","styleName","medium","exists",true,"shoppingCartId","shoppingCart",{"_12":71,"_72":73,"_74":37,"_75":76,"_77":78,"_79":80,"_81":37},"locale","en","readerTheme","light","constants",{"_69":70},"publishers",{"_65":66,"_67":68},"learnerOrganizations",[],"needsLocaleCookieSync",false,"announcements",{"_59":60,"_61":-5},"sentryEnvironment","production","sentryRelease","cce21d19a0d9612afdbced29947de1d8868e588b","announcement_store",{"_62":52,"_63":64},"announcement_authors","flag","text","

Kick off your book project, get started with GhostAI, get better at marketing, or spend the day doing all three! Free live workshops on Zoom. Saturday, June 27, 2026.

\n","companies",[],"universities",[],"paypalClientId","AX53Cwz8jR6rVVzzZiZNWWXl0dGagi-GZrjG-S3Ua14E5nrKMn7DxDQ730O4R8LusLPp49ALn6aRVaEy","no-cart","count",0,"free","updatedAt","","paymentGateway","STRIPE","cartItems",[],"loading"]

Email the Author